Quickstart

1. Get your endpoint and key

Value	What it is
`BASE_URL`	Your inference endpoint. Endpoints differ between accounts — use yours exactly as provided, without adding or stripping a path.
`MODEL`	The model your endpoint is provisioned for
`API_KEY`	Your account’s API key. Send it as a bearer token on every request.

You can create more keys under API Keys on the platform.

2. Install the SDK

Impala is OpenAI-compatible, so use the official openai package.

pip install openai

3. Send your first request

Fill in your three values and run it.

from openai import OpenAI

client = OpenAI(
    base_url="<BASE_URL>",
    api_key="<API_KEY>",
    timeout=600.0,
)

resp = client.chat.completions.create(
    model="<MODEL>",
    messages=[
        {"role": "system", "content": "You are a concise assistant."},
        {"role": "user", "content": "What is 17 times 3? Answer with just the number."},
    ],
)
print(resp.choices[0].message.content)

curl "<BASE_URL>/chat/completions" \
  -H "Authorization: Bearer <API_KEY>" \
  -H "Content-Type: application/json" \
  --max-time 600 \
  -d '{
    "model": "<MODEL>",
    "messages": [{"role": "user", "content": "What is 17 times 3?"}]
  }'

4. Check it worked

A response with no error means your endpoint, key and model are working end to end.

Chat completions

The request shape, streaming, timeouts and parameters.

Run a batch

File in, results out, at the lowest cost per token.

Bring Your Own Cloud installs differently — see Install Impala in your VPC.

Serverless inference

Batch inference

Agent orchestration

Optimization

Bring your own model

Bring your own cloud (BYOC)

Integrations

1. Get your endpoint and key

2. Install the SDK

3. Send your first request

4. Check it worked

Next

Chat completions

Run a batch

​1. Get your endpoint and key

​2. Install the SDK

​3. Send your first request

​4. Check it worked

​Next

Chat completions

Run a batch

1. Get your endpoint and key

2. Install the SDK

3. Send your first request

4. Check it worked

Next